Unifying Diffusion, Score-Based, and Flow Matching Models
A new measure-theoretic framework unifies diffusion, score-based, and flow matching generative models — the mathematical backbone of modern AI video and image synthesis systems.
GetReal Expands Deepfake Forensics and Identity Defense
GetReal Security is doubling down on deepfake detection with a new Forensics Spotlight capability and an RPO partnership aimed at scaling identity-security defenses against AI-generated impersonation attacks.
ML System Design: Data Labeling for Content Moderation
A deep dive into designing data labeling pipelines for content moderation systems—critical infrastructure for detecting harmful synthetic media, deepfakes, and policy-violating AI-generated content at scale.
Deepfake Fraud Forces Banks to Rebuild Defenses
Synthetic deepfakes are driving a wave of financial fraud, pushing banks to deploy new biometric and liveness detection systems as customer funds face escalating risk from voice cloning and face-swap attacks.
EU Agrees to Ban Non-Consensual Sexual Deepfake Tools
European Union lawmakers have agreed to prohibit AI tools designed to create non-consensual sexualised deepfakes, marking one of the most significant regulatory moves yet against synthetic media abuse.
Beijing AI Lab Z.ai Hits $20B as China Lures Investors
A Beijing-based AI lab is reportedly closing in on a $20 billion valuation as Western investors increasingly pivot toward Chinese frontier model developers amid surging competition with US labs.
Sumsub Launches Real-Time Deepfake Detection Engine
Identity verification firm Sumsub has rolled out a real-time deepfake detection system aimed at stopping live video-based fraud, signaling a shift from post-hoc analysis to in-session synthetic media defense.
SubQ: Miami Startup's Attention Trick Runs 52x Faster
A four-person Miami startup called SubQ claims a new attention mechanism that runs 52x faster than standard transformers at one-fifth the cost of Claude Opus, hinting at cheaper long-context inference for video and multimodal AI.
76% of UK Firms Hit by Deepfake Attacks, Few Prepared
New research finds 76% of UK organizations have faced deepfake attacks, yet most lack the detection tools, training, and incident response plans needed to defend against synthetic media threats targeting executives and finance teams.
Spotify Bets Big on AI-Generated Personal Audio
Spotify is positioning itself as the central distribution hub for AI-generated personal audio, signaling a major platform shift toward synthetic voice content and personalized listening experiences powered by generative AI.
Spotify AI DJ Expands Voice Cloning to 4 New Languages
Spotify's AI DJ now speaks French, German, Italian, and Brazilian Portuguese, leveraging OpenAI voice synthesis and Sonantic-derived cloning tech to deliver localized synthetic audio personalities to global users.
Detecting Human-LLM Co-Authored Text Boundaries
New research applies change point detection to identify exactly where human writing ends and LLM-generated text begins in co-authored documents — a critical advance for content authenticity verification.